Skip to content

Add DG_MINIMIZE_NUM_SMS env var to control SM minimization #273

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
yurekami wants to merge 1 commit into from
Closed

Add DG_MINIMIZE_NUM_SMS env var to control SM minimization #273

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 3 additions & 1 deletion csrc/jit_kernels/heuristics/sm100.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -100,7 +100,9 @@ struct SM100ArchSpec {
}

static bool should_minimize_num_sms() {
return true;
// Allow control via environment variable (default: enabled)
// Set DG_MINIMIZE_NUM_SMS=0 to disable SM minimization for more stable TTFT
return get_env<int>("DG_MINIMIZE_NUM_SMS", 1) != 0;
}

static std::pair<bool, bool> get_multicast_legality(const GemmType& gemm_type, const int& num_groups,
Expand Down
4 changes: 3 additions & 1 deletion csrc/jit_kernels/heuristics/sm90.hpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -99,7 +99,9 @@ struct SM90ArchSpec {
}

static bool should_minimize_num_sms() {
return true;
// Allow control via environment variable (default: enabled)
// Set DG_MINIMIZE_NUM_SMS=0 to disable SM minimization for more stable TTFT
return get_env<int>("DG_MINIMIZE_NUM_SMS", 1) != 0;
}

static std::pair<bool, bool> get_multicast_legality(const GemmType& gemm_type, const int& num_groups,
Expand Down